INDEX

Explanations

words related to negative or problematic situations, often involving mistakes or failures

oai_token-act-pair · gpt-3.5-turbo

negative expressions related to distress and adverse experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

izes

-0.89

ively

-0.80

acion

-0.73

eway

-0.73

ership

-0.71

abet

-0.71

inen

-0.70

ivity

-0.70

iveness

-0.69

iola

-0.69

POSITIVE LOGITS

nesday

0.93

uled

0.90

ĸļ

0.80

 awake

0.78

 away

0.77

out

0.76

GGGGGGGG

0.75

 onto

0.75

 aback

0.70

CLASSIFIED

0.69

Activations Density 0.174%

words related to negative or problematic situations, often involving mistakes or failures

negative expressions related to distress and adverse experiences

No Comments

No Known Activations