INDEX

Explanations

verbs that indicate struggle or difficulty

oai_token-act-pair · gpt-3.5-turbo

instances of struggle or difficulty in achieving various goals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Fax

-0.83

usra

-0.81

çīĪ

-0.77

prints

-0.72

 Tribune

-0.72

azon

-0.70

 Torch

-0.68

YES

-0.68

soType

-0.66

 Sign

-0.66

POSITIVE LOGITS

 cope

0.93

 recovering

0.89

 grasp

0.86

 reconcile

0.84

 regain

0.79

 recover

0.79

 rebound

0.79

 conceive

0.79

 fend

0.77

 reproduce

0.76

Activations Density 0.175%

verbs that indicate struggle or difficulty

instances of struggle or difficulty in achieving various goals

No Comments

No Known Activations