INDEX

Explanations

references to progress or improvement

oai_token-act-pair · gpt-3.5-turbo

references to progress

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

ENA

-0.73

 Peninsula

-0.67

 gorge

-0.64

 duck

-0.62

ski

-0.61

 resorts

-0.61

Natural

-0.60

FACE

-0.60

 cust

-0.59

stal

-0.59

POSITIVE LOGITS

ivism

1.40

iveness

1.12

ivity

1.09

ives

1.07

ions

1.06

ivist

1.03

 progress

0.95

ively

0.91

ments

0.90

 Progress

0.89

Activations Density 0.016%

references to progress or improvement

references to progress

No Comments

No Known Activations