INDEX

Explanations

phrases related to violence and conflict

oai_token-act-pair · gpt-3.5-turbo

terms associated with violence and firearms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

pend

-0.65

 Huntington

-0.63

 Negative

-0.61

hoe

-0.60

oat

-0.57

aughed

-0.56

CTR

-0.54

 concerned

-0.53

jun

-0.53

ãĥ£

-0.53

POSITIVE LOGITS

aciously

0.74

waves

0.70

ths

0.70

eper

0.68

EMA

0.67

tics

0.65

Film

0.64

steen

0.64

.>>

0.63

 with

0.63

Activations Density 0.188%

phrases related to violence and conflict

terms associated with violence and firearms

No Comments

No Known Activations