INDEX

Explanations

terms related to the concept of "extra"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "extra" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

theless

-0.83

iage

-0.74

better

-0.71

 Decay

-0.65

REDACTED

-0.63

ById

-0.63

 Cosponsors

-0.62

 Goodbye

-0.62

 CrossRef

-0.61

\\\\\\\\

-0.61

POSITIVE LOGITS

ordinary

1.52

judicial

1.51

verted

1.14

veh

1.05

ord

1.04

 virgin

0.91

verts

0.87

vas

0.87

ocular

0.84

 curric

0.84

Activations Density 0.029%

terms related to the concept of "extra"

instances of the word "extra" and its variations

No Comments

No Known Activations