INDEX

Explanations

fractions or ratios

oai_token-act-pair · gpt-3.5-turbo

phrases indicating proportions or ratios

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

irez

-0.61

andals

-0.60

Offline

-0.59

ymm

-0.59

idel

-0.56

isl

-0.55

ourgeois

-0.54

voice

-0.54

hai

-0.54

},"

-0.54

POSITIVE LOGITS

 every

0.90

ten

0.79

0.77

 tens

0.74

 equals

0.70

 bounds

0.69

0.68

0.67

Activations Density 0.032%

fractions or ratios

phrases indicating proportions or ratios

No Comments

No Known Activations