INDEX

Explanations

text related to descriptions or explanations

oai_token-act-pair · gpt-3.5-turbo

references to descriptions, outlines, and categories of information

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ococ

-0.82

ctors

-0.72

andom

-0.68

unique

-0.67

estial

-0.64

uries

-0.63

veland

-0.62

plete

-0.61

 lasted

-0.61

 purse

-0.60

POSITIVE LOGITS

 above

1.42

 earlier

1.13

above

1.06

 mentioned

1.05

mentioned

1.00

 supra

0.95

 outlined

0.94

 below

0.93

 alluded

0.92

 aforementioned

0.91

Activations Density 0.544%

text related to descriptions or explanations

references to descriptions, outlines, and categories of information

No Comments

No Known Activations