INDEX

Explanations

instances where information or details become known or come to light

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "surfaced."

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

ihad

-0.72

 dependence

-0.64

gone

-0.62

outer

-0.61

gee

-0.61

itar

-0.60

industrial

-0.60

jab

-0.60

aldi

-0.60

orate

-0.60

POSITIVE LOGITS

GROUND

0.82

Documents

0.72

 microphones

0.71

FACE

0.71

ãĤ¤ãĥĪ

0.70

 Weiner

0.70

BACK

0.69

âĸĳ

0.69

ãĥ¤

0.68

quished

0.67

Activations Density 0.040%

instances where information or details become known or come to light

occurrences of the word "surfaced."

No Comments

No Known Activations