INDEX

Explanations

studies, analysis, and surveys conducted by researchers

oai_token-act-pair · gpt-3.5-turbo

references to data analysis and research findings

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

atron

-0.82

ugal

-0.72

Assad

-0.71

inventoryQuantity

-0.70

rises

-0.68

Nib

-0.67

bec

-0.67

dyl

-0.67

Token

-0.67

Shift

-0.67

POSITIVE LOGITS

 participants

0.99

 randomized

0.97

 simulations

0.95

 undergraduate

0.89

 observational

0.86

 anonym

0.85

 cohorts

0.85

 genomic

0.85

 samples

0.85

 datasets

0.84

Activations Density 0.245%

studies, analysis, and surveys conducted by researchers

references to data analysis and research findings

No Comments

No Known Activations