INDEX

Explanations

mentions of the adjective "low" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to low-cost or low-level attributes

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

tnc

-0.86

andise

-0.85

itutional

-0.71

 Orient

-0.69

indal

-0.68

ophers

-0.68

 Ashe

-0.67

 Mahjong

-0.67

arium

-0.66

natureconservancy

-0.66

POSITIVE LOGITS

 enough

0.86

est

0.85

 hanging

0.83

 pitched

0.82

enthal

0.82

ball

0.79

ered

0.75

ppy

0.75

(<

0.74

ened

0.74

Activations Density 0.029%

mentions of the adjective "low" in various contexts

references to low-cost or low-level attributes

No Comments

No Known Activations