INDEX

Explanations

text related to listing items or steps in a sequence

oai_token-act-pair · gpt-3.5-turbo

references to lists or enumerations within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

utenant

-0.77

ilage

-0.75

uca

-0.74

acus

-0.72

raper

-0.71

ppe

-0.71

meier

-0.70

orescence

-0.70

ãĤ¨ãĥ«

-0.70

usk

-0.69

POSITIVE LOGITS

 paragraphs

1.16

 statements

1.04

 subsections

1.02

mentioned

1.02

 excerpts

1.00

 sections

0.98

 excerpt

0.96

 scenarios

0.95

 propositions

0.94

 phrases

0.92

Activations Density 0.073%

text related to listing items or steps in a sequence

references to lists or enumerations within the text

No Comments

No Known Activations