INDEX

Explanations

instances of the word 'that' followed by a description or related action

oai_token-act-pair · gpt-3.5-turbo

references to specific subjects or concepts that are being discussed or referenced

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

heny

-0.80

olics

-0.80

dan

-0.80

aughtered

-0.77

icons

-0.73

ysical

-0.73

oran

-0.73

eric

-0.72

lished

-0.71

cised

-0.71

POSITIVE LOGITS

 latter

0.99

 phenomenon

0.91

 assertion

0.90

 happening

0.88

 limitation

0.85

 happen

0.84

 trend

0.84

 sort

0.83

 complication

0.83

 notion

0.81

Activations Density 0.246%

instances of the word 'that' followed by a description or related action

references to specific subjects or concepts that are being discussed or referenced

No Comments

No Known Activations