INDEX

Explanations

instances of fear or potential dangers

oai_token-act-pair · gpt-3.5-turbo

expressions of fear related to various situations or consequences

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

•layer 5 - fear

No Comments

Negative Logits

Collins

-0.73

ascus

-0.73

ail

-0.69

ILLE

-0.69

moderate

-0.61

educ

-0.60

Norm

-0.60

flo

-0.60

Ni

-0.59

achel

-0.58

POSITIVE LOGITS

 fears

0.82

 reasons

0.79

 lack

0.77

ecause

0.77

 sheer

0.74

 proximity

0.69

 circumstances

0.69

 concerns

0.69

 feared

0.66

 cumbers

0.66

Activations Density 0.704%

instances of fear or potential dangers

expressions of fear related to various situations or consequences

No Comments

No Known Activations