INDEX

Explanations

words related to conflicts or battles

oai_token-act-pair · gpt-3.5-turbo

instances of the word "fighting" in various contexts

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

OGR

-0.71

uras

-0.69

endon

-0.69

gow

-0.68

 Atmospheric

-0.68

ographical

-0.67

ical

-0.66

ISTER

-0.65

ICAN

-0.65

ħĭ

-0.64

POSITIVE LOGITS

fighting

1.22

fights

0.93

 waged

0.92

fighters

0.90

fight

0.89

 raged

0.87

 against

0.83

fighter

0.78

 fought

0.78

 fatig

0.76

Activations Density 0.028%

words related to conflicts or battles

instances of the word "fighting" in various contexts

No Comments

No Known Activations