INDEX

Explanations

events or incidents being discussed or happening in various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "events" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

zzi

-0.75

geon

-0.74

acked

-0.70

sburgh

-0.70

Äĩ

-0.68

sis

-0.68

 Canal

-0.66

brid

-0.65

bal

-0.63

redo

-0.63

POSITIVE LOGITS

hips

1.06

 events

0.97

 unfold

0.89

 unfolded

0.87

etter

0.86

 transpired

0.86

ourcing

0.84

hip

0.80

ivals

0.79

 unfolding

0.79

Activations Density 0.017%

events or incidents being discussed or happening in various contexts

instances of the word "events" and its variations

No Comments

No Known Activations