INDEX

Explanations

mentions of the word or concept "adapt"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "adapt."

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

$$$$

-0.91

holder

-0.81

++++++++++++++++

-0.75

FORE

-0.73

Ranked

-0.70

gun

-0.70

 Roads

-0.69

Jr

-0.68

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.67

########

-0.67

POSITIVE LOGITS

 adapted

0.88

ively

0.88

 Adapt

0.86

 adapt

0.83

 adapting

0.82

adapt

0.82

etically

0.81

ability

0.79

 adaptation

0.77

atile

0.75

Activations Density 0.024%

mentions of the word or concept "adapt"

instances of the word "adapt."

No Comments

No Known Activations