INDEX

Explanations

names, possibly of individuals or locations

oai_token-act-pair · gpt-3.5-turbo

names of specific individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥĩãĤ£

-0.74

ected

-0.67

ective

-0.65

birth

-0.64

fax

-0.64

 vanquished

-0.63

 undone

-0.63

ioch

-0.61

 dividing

-0.60

ization

-0.59

POSITIVE LOGITS

atoon

0.87

lio

0.84

matically

0.82

Nug

0.81

itionally

0.80

uggets

0.80

uably

0.78

lette

0.77

untu

0.74

rats

0.74

Activations Density 0.016%

names, possibly of individuals or locations

names of specific individuals

No Comments

No Known Activations