INDEX

Explanations

words within a dialogue often followed by quotation marks such as 'I,' 'said,' or 'asked.'

oai_token-act-pair · gpt-3.5-turbo

instances of characters speaking or dialogue in conversations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 confir

-0.70

 fres

-0.69

 undermin

-0.67

 identity

-0.64

utterstock

-0.63

 wedd

-0.63

 representation

-0.63

oÄŁ

-0.62

 moder

-0.61

 helicop

-0.60

POSITIVE LOGITS

ÂłÂłÂłÂłÂłÂłÂłÂł

1.08

ÂłÂłÂł

1.05

ÂłÂłÂłÂł

1.01

"...

0.96

"âĢ¦

0.92

::::::::

0.89

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.85

"-

0.84

xxxxxxxx

0.83

Suddenly

0.82

Activations Density 0.123%

words within a dialogue often followed by quotation marks such as 'I,' 'said,' or 'asked.'

instances of characters speaking or dialogue in conversations

No Comments

No Known Activations