INDEX

Explanations

mentions of the word "dock" in texts

oai_token-act-pair · gpt-3.5-turbo

references to docks and docking processes

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

 Values

-0.67

 Isis

-0.67

char

-0.66

 Spears

-0.66

 Ezekiel

-0.66

 Charisma

-0.64

Ep

-0.63

 Kelley

-0.63

KR

-0.61

 values

-0.61

POSITIVE LOGITS

 dock

4.19

 docks

3.07

 Dock

2.80

 docking

2.58

 waterfront

1.07

 sail

1.06

 ashore

1.06

 lighthouse

1.04

 ferry

1.04

bay

1.01

Activations Density 0.010%

mentions of the word "dock" in texts

references to docks and docking processes

No Comments

No Known Activations