INDEX

Explanations

words related to non-fiction literature

oai_token-act-pair · gpt-3.5-turbo

references to different genres of fiction and non-fiction writing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

vis

-0.77

bat

-0.65

Roh

-0.64

 Bundes

-0.63

pac

-0.63

Wyn

-0.63

tip

-0.62

atz

-0.62

 adren

-0.61

 crown

-0.61

POSITIVE LOGITS

fiction

3.52

 Fiction

2.22

Ð´

1.13

ambling

1.00

translation

0.92

flu

0.91

obook

0.89

bender

0.89

 Drama

0.87

atche

0.82

Activations Density 0.037%

words related to non-fiction literature

references to different genres of fiction and non-fiction writing

No Comments

No Known Activations