INDEX

Explanations

words related to ghosts

oai_token-act-pair · gpt-3.5-turbo

mentions of ghosts

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

unes

-0.79

tics

-0.77

erity

-0.76

ibaba

-0.74

NRS

-0.74

ECA

-0.73

onian

-0.72

ahime

-0.72

UTERS

-0.71

thritis

-0.70

POSITIVE LOGITS

busters

1.10

writer

0.95

writing

0.92

buster

0.90

written

0.88

 haunting

0.87

glass

0.83

door

0.80

ghost

0.80

writers

0.78

Activations Density 0.020%

words related to ghosts

mentions of ghosts

No Comments

No Known Activations