INDEX

Explanations

words related to determining or attributing causation or responsibility

oai_token-act-pair · gpt-3.5-turbo

various forms and aspects of the concept of "determinants" and "contributions."

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

 Penet

-0.65

RESULTS

-0.64

 Surgery

-0.63

 motto

-0.62

 Valent

-0.61

bye

-0.61

NAV

-0.61

 Parish

-0.60

 Sisters

-0.60

 Masters

-0.60

POSITIVE LOGITS

itives

1.24

orship

1.13

ements

1.12

ctive

1.11

uable

1.09

ants

1.09

itional

1.09

ations

1.07

ates

1.04

ators

1.04

Activations Density 0.036%

words related to determining or attributing causation or responsibility

various forms and aspects of the concept of "determinants" and "contributions."

No Comments

No Known Activations