INDEX

Explanations

phrases related to presenting information or reporting facts

oai_token-act-pair · gpt-3.5-turbo

references to research, studies, or academic analysis

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

$$$$

-0.81

 Airl

-0.71

akable

-0.64

}}

-0.64

icide

-0.64

edes

-0.61

qt

-0.60

ideshow

-0.60

icides

-0.59

]."

-0.59

POSITIVE LOGITS

 gist

1.14

Specifically

0.91

Firstly

0.88

 nutshell

0.86

 excerpt

0.86

 premise

0.86

 Topics

0.83

 quote

0.82

 first

0.82

itled

0.81

Activations Density 0.728%

phrases related to presenting information or reporting facts

references to research, studies, or academic analysis

No Comments

No Known Activations