INDEX

Explanations

terms related to comparisons between different entities or categories

oai_token-act-pair · gpt-3.5-turbo

phrases that denote comparisons or contrasts generally indicating levels of inequality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

akeru

-0.86

istries

-0.85

arios

-0.76

liga

-0.73

 ãĤµ

-0.62

 thanking

-0.62

ONEY

-0.61

 Lisp

-0.60

âĢİ

-0.58

 Consent

-0.58

POSITIVE LOGITS

 slightest

0.73

 ones

0.70

 anymore

0.70

outer

0.69

 predecessors

0.68

 rivals

0.67

 counterparts

0.66

 altogether

0.65

 itself

0.65

 equals

0.64

Activations Density 0.492%

terms related to comparisons between different entities or categories

phrases that denote comparisons or contrasts generally indicating levels of inequality

No Comments

No Known Activations