INDEX

Explanations

numeric or word representations of scores or values

oai_token-act-pair · gpt-3.5-turbo

references to real-time data and its applications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 ILCS

-0.82

 Strait

-0.77

ynski

-0.77

 Rica

-0.72

ulhu

-0.70

 Norris

-0.69

AVG

-0.69

kson

-0.67

 Cotton

-0.66

oglu

-0.66

POSITIVE LOGITS

estate

1.11

element

0.99

science

0.97

time

0.95

naissance

0.94

historic

0.93

terms

0.92

condition

0.89

sounding

0.89

repeat

0.88

Activations Density 0.046%

numeric or word representations of scores or values

references to real-time data and its applications

No Comments

No Known Activations