INDEX

Explanations

words related to physical locations or sites

oai_token-act-pair · gpt-3.5-turbo

references to specific locations or sites

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

olyn

-0.79

sen

-0.75

lass

-0.74

 Lauder

-0.71

 Franch

-0.71

atures

-0.68

ornia

-0.68

 outgoing

-0.68

 ILCS

-0.67

ths

-0.64

POSITIVE LOGITS

bench

0.84

kit

0.77

geist

0.76

ifier

0.75

ologies

0.75

keeper

0.73

 sites

0.71

 administrator

0.71

osphere

0.71

 admins

0.69

Activations Density 0.037%

words related to physical locations or sites

references to specific locations or sites

No Comments

No Known Activations