INDEX

Explanations

phrases related to measurement and statistics

oai_token-act-pair · gpt-3.5-turbo

quantitative information about actions and incidents

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

odox

-0.92

çīĪ

-0.85

aroo

-0.84

eland

-0.74

sonian

-0.73

hon

-0.72

ican

-0.72

kat

-0.71

idon

-0.70

icol

-0.69

POSITIVE LOGITS

 Median

0.94

(%

0.94

 Percentage

0.92

Avg

0.91

 Average

0.85

Average

0.84

 households

0.81

per

0.79

 accounted

0.79

 median

0.77

Activations Density 0.580%

phrases related to measurement and statistics

quantitative information about actions and incidents

No Comments

No Known Activations