INDEX

Explanations

words related to an inner aspect or hidden part

oai_token-act-pair · gpt-3.5-turbo

the term "inner" as it relates to various contexts, particularly urban or hidden dynamics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

UTC

-0.70

Mario

-0.67

Canada

-0.67

ATURES

-0.67

je

-0.67

Trend

-0.66

bill

-0.65

 Morales

-0.64

cert

-0.62

 Stat

-0.62

POSITIVE LOGITS

 inner

3.76

 Inner

2.42

 outer

2.19

inner

1.68

 Outer

1.51

 inward

1.49

 internal

1.42

 interior

1.42

internal

1.27

outer

1.26

Activations Density 0.018%

words related to an inner aspect or hidden part

the term "inner" as it relates to various contexts, particularly urban or hidden dynamics

No Comments

No Known Activations