INDEX

Explanations

mentions of high temperature or warmth

oai_token-act-pair · gpt-3.5-turbo

references to heat or temperature-related subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

VICE

-0.82

 Mandela

-0.70

 Goff

-0.69

âĢ¢âĢ¢

-0.69

 Merry

-0.67

allery

-0.67

ournal

-0.64

RB

-0.64

ufact

-0.63

 Lynd

-0.62

POSITIVE LOGITS

hens

1.36

seekers

1.11

waves

1.05

wave

1.04

hen

0.92

stroke

0.89

 oven

0.83

Heat

0.83

 exhaustion

0.82

 exch

0.79

Activations Density 0.033%

mentions of high temperature or warmth

references to heat or temperature-related subjects

No Comments

No Known Activations