INDEX

Explanations

phrases or words indicating an exception or an alternative

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate exceptions or additional information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 plurality

-0.57

rye

-0.56

 tumble

-0.53

 disinfect

-0.53

 machine

-0.52

 Meadow

-0.51

 shutter

-0.51

 deposition

-0.51

 Pace

-0.51

ettle

-0.51

POSITIVE LOGITS

heid

1.22

ments

1.07

icularly

0.91

ranging

0.88

selves

0.85

entimes

0.85

ional

0.84

rompt

0.80

ractor

0.80

ively

0.78

Activations Density 0.033%

phrases or words indicating an exception or an alternative

phrases that indicate exceptions or additional information

No Comments

No Known Activations