INDEX

Explanations

words related to physical or verbal threats or coercion

oai_token-act-pair · gpt-3.5-turbo

instances of intimidation and related actions

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

ports

-0.80

Foot

-0.79

izen

-0.75

noon

-0.74

launch

-0.73

empl

-0.72

argo

-0.71

odes

-0.71

night

-0.70

 Fault

-0.69

POSITIVE LOGITS

 intimidation

1.05

 tactics

0.97

 intimidated

0.96

 intimid

0.87

 intimidate

0.87

 coer

0.86

 stalking

0.82

 retaliation

0.78

 tactic

0.78

 blackmail

0.75

Activations Density 0.028%

words related to physical or verbal threats or coercion

instances of intimidation and related actions

No Comments

No Known Activations